Skip to Main content Skip to Navigation
Reports

Budget-aware workflow scheduling with DIET

Aurélie Kong Win Chang 1, 2 Yves Caniou 1 Eddy Caron 1 Yves Robert 2, 3
1 AVALON - Algorithms and Software Architectures for Distributed and HPC Platforms
Inria Grenoble - Rhône-Alpes, LIP - Laboratoire de l'Informatique du Parallélisme
2 ROMA - Optimisation des ressources : modèles, algorithmes et ordonnancement
Inria Grenoble - Rhône-Alpes, LIP - Laboratoire de l'Informatique du Parallélisme
Abstract : DIET is a widely used workflow engine capable of scheduling workflows dynamically. This feature enables the engine to react to various events occurring during execution. However, a static schedule calculated before the submission of the workflow can provide many optimization opportunities, in particular when budget constraints should be enforced. In this paper, we expand DIET by adding a static scheduling functionality. To validate this new functionality on a realistic test case, we compare the behavior of budgetaware algorithms designed in [5], both in simulation and on an actual platform, Grid’5000. The obtained results are very similar, showing that the benefits provided by the scheduling algorithms are effective, both in simulation and in a real execution.
Document type :
Reports
Complete list of metadatas

https://hal.inria.fr/hal-03080468
Contributor : Equipe Roma <>
Submitted on : Friday, December 18, 2020 - 10:24:35 AM
Last modification on : Saturday, December 19, 2020 - 3:40:06 AM

File

RR-9381.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-03080468, version 1

Collections

Citation

Aurélie Kong Win Chang, Yves Caniou, Eddy Caron, Yves Robert. Budget-aware workflow scheduling with DIET. [Research Report] RR-9381, Inria Grenoble Rhône-Alpes. 2020. ⟨hal-03080468⟩

Share

Metrics

Record views

29

Files downloads

104